Animal behavior depends on the function of a large collection of overlapping neural circuits. To fully under- stand the circuit underlying a particular behavior, one must identify the neurons involved, determine what synaptic connections they make with each other, and measure their electrical responses during activation of the circuit. The zebrafish larva is an excellent system to study circuits: it has well-established behaviors, can be manipulated genetically, and most importantly, is transparent. By genetically expressing fluorescent reporters or light-activated channels, one can optically image neurons' morphology, connectivity, and activity, and even optically control their electrical activity, in an intact, living animal (Scott, 2009). What has been largely missing, until recently, are methods to express genes in particular neurons of interest. A powerful solution to this problem is provided by Gal4 enhancer trap screens in zebrafish (Scott et al., 2007; Asakawa et al., 2008). The Gal4 gene, which acts as a genetic trigger, is integrated randomly into the zebrafish genome; depending on where it lands, it will be turned on in a different set of cells (often including specific neuronal types), controlled by the regulatory elements of nearby genes. By screening through many Gal4 mutant lines, one can find lines that express in one's favorite neurons, then cross these to UAS responder lines, so that fluorescent reporters or other genes are turned on in those neurons. This proposal will carry out a second-generation Gal4 enhancer trap screen with several improvements. (1) A new DNA trapping construct that not only expresses Gal4, but can be converted to instead express a different genetic switch, Cre recombinase. This will allow expression of genes in even more specific sets of cells by intersecting a Gal4 pattern with a Cre pattern. (2) An online database of Gal4 expression patterns, including three-dimensional views. This will allow collaborators, and eventually the zebrafish community at large, to quickly determine which lines may express in the tissues or neurons that they study. (3) A new public- domain 3D visualization package, FluoRender, that has been optimized for confocal microscopy data. This will improve and speed up documentation of expression patterns. (4) A toolkit of UAS responder lines, validated for uniform expression levels, to visualize neuronal shape and connectivity. Diencephalic dopaminergic neurons, for which no specific enhancer is yet known, will be analyzed as a test case. In summary, then, this project will generate a large number of well-characterized Gal4 enhancer trap lines and UAS responder lines, which will allow zebrafish neurobiologists as well as other zebrafish researchers to express genes of interest specifically in many different neuron classes and nonneural tissues. Techniques developed for intersectional gene expression, generation of UAS responders, and 3D visualization will also be widely applicable in the field. The project will significantly increase the utility of the Gal4-UAS method in zebrafish, aiding analysis of the development and function of many organs, in addition to neuronal circuits.